• Please review our updated Terms and Rules here

DEC RX02 Diskettes

BitWiz

BitWiz

Experienced Member
Joined
Sep 7, 2021
Messages
420
Location
Glen Ellyn, Iil
I have about 2 dozen DEC RX02 8" Diskettes. These are the DEC branded floppies.

Almost all of them have 1 or more bad sectors (determined by using the /R option of RXCOPY after formatting with the /D option).

I have tried bulk erasing them, track erasing them in a greaseweazle, force formatting them in the greaseweazle and nothing seems to work.

Why would the DEC diskettes be so bad vs the Verbatim, Dysan, BASF and 3M diskettes that I have?

Also, does OS/8 do any kind of bad block mapping on RX02s like they do on the RK05?
 
I recently went through about 100 of those earlier this year--all DEC-branded RX02, but with two different styles of manufacturer's labels. So DEC probably used two different suppliers. FWIW, none had unrecoverable errors (RX01 formatted), but one had lots of dust and crud (probably stored without the jacket on top of someone's bookcase). I had to give that one a new jacket after cleaning.
One batch was the usual drab brown-and white label; the other has the blue and white label in the corner saying "Digital disk"brown.pngblue.png


The drive I used to read them was a Siemens FDD-200.
 
BitWiz said:
Also, does OS/8 do any kind of bad block mapping on RX02s like they do on the RK05?
Click to expand...
No it doesn't. Didn't think RK05's did either. I know RL0x do bad block mapping.

Did you inspect the surface for dirt/mold spots? Could it be poor storage?

I haven't had particular problems with DEC branded disks though don't have a huge number of them.
 
Chuck(G) said:
I recently went through about 100 of those earlier this year--all DEC-branded RX02, but with two different styles of manufacturer's labels. So DEC probably used two different suppliers. FWIW, none had unrecoverable errors (RX01 formatted), but one had lots of dust and crud (probably stored without the jacket on top of someone's bookcase). I had to give that one a new jacket after cleaning.
One batch was the usual drab brown-and white label; the other has the blue and white label in the corner saying "Digital disk"View attachment 1263198View attachment 1263199


The drive I used to read them was a Siemens FDD-200.
Click to expand...
All but one I have look like the one on the left and almost all of them cannot be formatted without at least one read error. The process is as follows:

The greaseweazle has two Shugart SA-800 drives that have successfully created bootable RX02 floppies on non-DEC diskettes.

Here is the process I used after a few failed. Please note the disks look brand new.

1. Bulk erase the disks on both sides.
2. Erase the disks in the Greastweazle
3. Attempt for format the disks in the Greasweazle - failure
4. On an RX02 Drive under OS/8:
.R RXCOPY
*RXA0:/D
5. On an RX02 Drive under OS/8:
.R RXCOPY
*RXA0:/R
This step produces multiple sector read errors.

The one disk that I have that matches the disk on the right formatted flawlessly.
 
djg said:
No it doesn't. Didn't think RK05's did either. I know RL0x do bad block mapping.

Did you inspect the surface for dirt/mold spots? Could it be poor storage?

I haven't had particular problems with DEC branded disks though don't have a huge number of them.
Click to expand...
You are correct, I was remembering the RL01 and not the RK05 as I stated.
 
Since a number of the disks that i can cannot be used. Does anyone have some spare single sided double density capable disks they would like to find a good home for?

Thanks everyone.
 
Chuck(G) said:
Have you tried formatting using a real FDC, not a Greaseweazle? Aren't RX01 media the same as RX02?
Click to expand...
I don't have anything other than the greaseweazle and the actual RX02.

I have tried taking a bulk erased disk directly to the RX02 and it won't format it BUT if I format track zero with the greaseweazle then the RX02 can format it with RXCOPY. Which is something that I didn't expect.

The only difference between the RX01 and the RX02 is that the RX02 writes the data portion of the sectors in double density but the sector header information is in single density and there is a bit in the header that says whether the data is single or double density.

From a hardware standpoint the drives are identical (that I'm aware of) but the drive controller board is different and I think the drive control board is different or at least as different firmware on it.
 
I'm very familiar with the RX02 (wacky MFM) format. Perhaps the RX02 drive can't format single-density (T0 is single-density; oddly, a lot of PDP8 software doesn't use it in any case).
 
I'm sorry to imply that you didn't know the RX02 specifics.

However, the OS/8 RXCOPY command can format single density (RX01) disks with the /S option or double density (RX02) with the / D option.

I have tried both formats with fewer bad sectors single density but there are still bad sectors.
 
But can RXCOPY format single-density sectors? Or does it simply attempt to read track 0 and die if it's not formatted? I've certainly seen RX01 floppies where the original track 0 is the standard IBM 3740 one, complete with EBCDIC label fields and deleted data sectors. I've also seen RX01 floppies with ordinary E5s on track 0.
 
Chuck(G) said:
But can RXCOPY format single-density sectors? Or does it simply attempt to read track 0 and die if it's not formatted? I've certainly seen RX01 floppies where the original track 0 is the standard IBM 3740 one, complete with EBCDIC label fields and deleted data sectors. I've also seen RX01 floppies with ordinary E5s on track 0.
Click to expand...
The documentation says that it will format single and double sensity sectors. However, that might just be setting the density bit in the header.

I know it can format all of the tracks except track zero. I will try to find the source code to it and see if I can modify it.
 
Actually, RXCOPY can't format at all. If the single density sector headers are not present then the diskette is unusable. You have to put the sector headers on the diskette using some non DEC drive and controller. What RXCOPY can do is write sector data after it finds the correct sector header. The /D or /S options probably force a write of every sector in either single or double density. It does a blind write. The /R probably does a read of every sector and if the checksum doesn't match that is an error. The /D would be useful for turning an RX01 diskette into an RX02 diskette. /S would be useful for turning an RX02 diskette into an RX01 diskette.

If you are having problem with media it could be bad media, bad head amps or could even be an alignment issue. Since RX0? never writes the sector headers and you can't format on those drives there is no way to tell for certain without an alignment diskette. If the media won't format on some other drive that otherwise works, then it is probably the media.
 
DougIngraham said:
Actually, RXCOPY can't format at all. If the single density sector headers are not present then the diskette is unusable. You have to put the sector headers on the diskette using some non DEC drive and controller. What RXCOPY can do is write sector data after it finds the correct sector header. The /D or /S options probably force a write of every sector in either single or double density. It does a blind write. The /R probably does a read of every sector and if the checksum doesn't match that is an error. The /D would be useful for turning an RX01 diskette into an RX02 diskette. /S would be useful for turning an RX02 diskette into an RX01 diskette.

If you are having problem with media it could be bad media, bad head amps or could even be an alignment issue. Since RX0? never writes the sector headers and you can't format on those drives there is no way to tell for certain without an alignment diskette. If the media won't format on some other drive that otherwise works, then it is probably the media.
Click to expand...
Since the ones that fail in the RX02/RX28 also fail in the Greaseweazle/SA-800 I will assume the media is bad. My original question was why is the DEC media so much worse than the Dysan/3M/BASF/Nashua disks that I have.

I am seeing a failure rate of greater than 50% on the DEC RX02K diskettes (DEC's Number). I also have a handful of disks that were "flippied" when I received them. Some of them are bad on the front but usable on the back.

A major issue with the OS/8 disk structure on the RX02 (and the RK05) is that there is no support for bad sectors. I've worked with different schemes where a bad sector can be mapped out or at worst a bad track. Any diskette that doesn't format and read correctly 100% is basically useless for OS/8.

I have successfully bulk erased a disk, format it with the greastweazle and then "reformatted" it in the RX02.

If memory serves there is a bit in the sector header that specifies the density of the sector data. The RXCOPY /D option sets this bit and writes a sector of data in double density (I am assuming this is what /D does from the sector data stand point) and the RXCOPY /S option clears this bit and writes the sector single density (again an assumption based on observation).
 
BitWiz said:
A major issue with the OS/8 disk structure on the RX02 (and the RK05) is that there is no support for bad sectors. I've worked with different schemes where a bad sector can be mapped out or at worst a bad track. Any diskette that doesn't format and read correctly 100% is basically useless for OS/8.
Click to expand...

Unless the bad block mapping stuff is in the controller, there is no way to make this work in an OS/8 handler. You get 93 words in a 1 page SYS handler and 221 in a 2 page handler. This is not enough to hold the table

One thing that can be done is to create a file BBXXX that points at the bad block. You can't tolerate a bad block in the system section or the directory of an OS/8 file system. with this approach.

BitWiz said:
I have successfully bulk erased a disk, format it with the greastweazle and then "reformatted" it in the RX02.
Click to expand...

The format was not done by the RX02. The RX02 just writes the double density sectors.

BitWiz said:
If memory serves there is a bit in the sector header that specifies the density of the sector data. The RXCOPY /D option sets this bit and writes a sector of data in double density (I am assuming this is what /D does from the sector data stand point) and the RXCOPY /S option clears this bit and writes the sector single density (again an assumption based on observation).
Click to expand...
I don't think the controller touches the sector headers.
 
DougIngraham said:
Unless the bad block mapping stuff is in the controller, there is no way to make this work in an OS/8 handler. You get 93 words in a 1 page SYS handler and 221 in a 2 page handler. This is not enough to hold the table

One thing that can be done is to create a file BBXXX that points at the bad block. You can't tolerate a bad block in the system section or the directory of an OS/8 file system. with this approach.



The format was not done by the RX02. The RX02 just writes the double density sectors.


I don't think the controller touches the sector headers.
Click to expand...
Doug,

You are correct, the density indicator is actually a byte the comes before the actual data sector but in the data sector and not the sector header.

Please accept my apology
 
You must log in or register to reply here.
Back
Top